Geothermal Exploration of Jura Canton (Switzerland) using Ambient Noise Tomography: Velocity and Attenuation models

Deep geothermal energy is expected to play a key role as a baseload resource within the Net Zero climate initiatives of the European Union and Switzerland. In Switzerland, the Haute Sorne project, led by Geo-Energie Suisse, represents the first Enhanced Geothermal System (EGS) initiative since the suspension of the Basel project in 2010. Approved by the Canton of Jura in January 2022, this project aims to develop and validate advanced exploration and monitoring technologies to harness the geothermal potential of Switzerland while minimising seismic risk.To reduce subsurface uncertainty, particularly at basement depths, we deployed a seismic network comprising 700 3C nodal sensors around the Haute Sorne area, with a radius of 12 kilometres. Starting in February 2024, these sensors continuously recorded 3-component ambient noise data for one month.

In this study, we applied the classical Ambient Noise Tomography (ANT) technique and the Ambient Noise Attenuation Tomography (ANAT) method and compared the resulting 3D shear-wave velocity and intrinsic attenuation models. Standard data processing techniques were employed to retrieve Empirical Green's Functions (EGFs) from ambient noise cross-correlations for Rayleigh and Love waves. The methodology was then divided into two ways: ANT and ANAT.

For ANT, dispersion curves were extracted using the FTAN (Frequency Time Analysis) technique. Group velocity maps for various periods were obtained for both Rayleigh and Love waves through a linearized inversion approach. A joint inversion of Rayleigh and Love dispersion curves was then conducted using a transdimensional Bayesian formulation to obtain a 3D shear-wave velocity model.

For ANAT we used the methodology described by Cabrera-Pérez et al. (2024). The intrinsic attenuation for each EGF was calculated across multiple frequencies for Rayleigh and Love waves using the lapse-time dependence method, which evaluates attenuation based on the coda window length at different onsets of the ambient noise cross-correlation coda. Next, 2D intrinsic attenuation maps for various frequencies were derived via linear inversion using sensitivity kernels. Finally, a joint inversion of Rayleigh and Love attenuation data was performed to obtain a 3D intrinsic attenuation model.

Our research seeks to map regional geological features at depths of up to 5 kilometres to improve seismic hazard assessments and gain a more comprehensive understanding of the local seismotectonic context. Particular attention is being focused on basement formations, including a suspected Permo-Carboniferous trough. By identifying variations in shear-wave velocity and attenuation, we aim to characterise the subsurface structure of this region. We find that these passive seismic exploration methods are particularly valuable when surveying regions with significant lateral variations in subsurface structure and topography and where classical seismic reflection campaigns may be logistically challenging and cost prohibitive.

References:
Cabrera-Pérez, I., D’Auria, L., Soubestre, J., Del Pezzo, E., Prudencio, J., Ibáñez, J. M., ... & Pérez, N. M. (2024). 3-D intrinsic attenuation tomography using ambient seismic noise applied to La Palma Island (Canary Islands). Scientific Reports, 14(1), 27354.